Post-marketing surveillance refers to any non-experimental or observational study, method, or monitoring strategy that is applied to obtain information on drug experience (primarily adverse) after a drug has been approved for clinical use. One of the major problems in post-marketing surveillance studies is the lack or under-reporting of drug experiences by health care professionals. This study was developed to describe the impact of three different prescription event monitoring programmes on the reporting of adverse drug reactions (ADR's) in the hospital situation. The intensive ADR monitoring programme and two voluntary ADR monitoring programmes which followed were conducted in the medical wards of an urban teaching and referral hospital. All patients admitted to the designated wards were monitored by a dedicated pharmacist in the intensive programme, ward pharmacists in the first voluntary programme and by medical and nursing staff in the second voluntary programme. The pharmacist monitored a cohort of patients prospectively in two medical wards for a period of three months. The patient's record was linked with any suspected ADR. All details, i.e. patient drug orders, characteristics and ADR description, were recorded and then reported. From 228 patients monitored, 25 cases have been reported. The impact of the intensive ADR monitoring programme was a reporting rate of 11 percent. Reports were received on ADR's of a particularly mild, common and pharmacologically predictable (type A) nature. The first voluntary ADR monitoring programme comprised the reporting of suspected AD R's and the recording of drug orders for the patients and the patient characteristics. The ward pharmacists monitored for suspected AD R's in all patients during their regular ward rounds. Six cases were reported in a population of 1506 patients monitored during the three months. The reports were mainly on moderate to severe suspected AD R's of pharmacologically unpredictable (type B) nature. The rate of reports received by the surveillance unit in this study was 4 reports per ward pharmacist per annum. The second voluntary ADR monitoring programme comprised the prospective monitoring of 1555 patients by medical and nursing staff during their stay at the designated medical wards during the three month period. Patients were monitored for any ADR and when an ADR was suspected, the patient characteristics and drug orders were recorded and reported to the surveillance unit. Ten cases were reported represented by six reports from doctors and four by sisters. The reporting rate was 2 reports per doctor in four years and 3 reports for each member of the nursing team in 5 years. Reports were mainly received on moderate to severe suspected ADR's of a pharmacologically unpredictable (type B) nature.